The long range goal of this research program is to construct a quantitative model of the immune system. This model will provide a means of analyzing with numerical and statistical precision the detailed logic that links genes to cells and cells to the immune response. Our research plan depends on disassembling the immune system into smaller and smaller logical units (e.g. B cells, humoral antibodies which act as receptors and effectors, immunoglobulin light and heavy chains, exons that encode various segments of these L and H chains, and even nucleotides that encode amino acids which determine antibody specificity). At each stage the logic of disassembly is used to reverse the process and reassemble the units into an immune system. The power of methodology is illustrated by its application to the humoral immune response. The following conclusions run counter to the conventional wisdom and highlight how easily one can be misled by intuition; the antibody repertoire contains a total of 6.7X10 4 functionally different specificities; only 10% of the total B cell population can be induced by antigen to secrete antibody; there can be no more than 0 different V exons (VL=VH=90) in the germline; and immunoglobulin receptors cannot function by antigen induced aggregation. Three areas need further study. First, although our model of B cell immunity has been constructed with careful attention to detail, one further test is needed to determine whether under the conditions of the model a single VL VH pair could have evolved to the subtle system found today in species as diverse as tadpoles and elephants. This testing of models against evolutionary forces is a key ingredient in the purely theoretical evaluation of the correctness of our models. Of course making and testing experimental predictions remains essential. Second, the methodology which has been successfully used to analyze the humoral immune system must now be applied to the cell-mediated immune response. The important features of the T cell compartment has been loosely outlined. Third, the skeleton of a computerized interactive immune response has been formulated and needs to be refined to incorporate the detailed component outputs of the humoral and cell mediated immune response compartments.